feat: expose 'runs-on' to container-cicd workflow

[ezdac]

Since we now run a lot of workflows in our self-hosted runners, there is a need to expose the runs-on option to the container-cicd workflow. The default is left at "ubuntu-latest" to avoid any compatibility issues, but it could
be overwritten with e.g. '["self-hosted", "org", "8-cpu"]'.

[pputman-clabs]

I'm trying to remember why we did this, but I think there was a problem with the caching related to using self hosted runners sometimes and local runners at other times, and they couldn't reference each others cache? I dont remember. I'm approving this for now but just be aware there may be a need to either revert this or make further changes.

[jcortejoso]

I'm trying to remember why we did this, but I think there was a problem with the caching related to using self hosted runners sometimes and local runners at other times, and they couldn't reference each others cache? I dont remember. I'm approving this for now but just be aware there may be a need to either revert this or make further changes.

I think it should work normally with independence of the node used, but maybe I'm ignoring something.

I'd just merge it.

[pputman-clabs]

I think its because the path that the cache is stored in, is different on self hosted vs github hosted, so it was breaking the cache if the same repo used one, then used the other. so if we're going to use a cache, the best thing to do is use the same type of runner for the same reusable workflow in the same repo, or find some way to configure the cache path based on if its privately or github hosted

…

On Tue, Sep 23, 2025 at 7:58 AM Javier Cortejoso ***@***.***> wrote: *jcortejoso* left a comment (celo-org/reusable-workflows#142) <#142 (comment)> I'm trying to remember why we did this, but I think there was a problem with the caching related to using self hosted runners sometimes and local runners at other times, and they couldn't reference each others cache? I dont remember. I'm approving this for now but just be aware there may be a need to either revert this or make further changes. I think it should work normally with independence of the node used, but maybe I'm ignoring something. I'd just merge it. — Reply to this email directly, view it on GitHub <#142 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AX2F4DSIIHDHYE2AL5GBZZL3UE7WPAVCNFSM6AAAAACDZANGW2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTGMRTHEYDOOBWGA> . You are receiving this because you commented.Message ID: ***@***.***>